I still don't know why - or even if - the Quebec labour market is doing as badly as the numbers I've presented here and in Economy Lab suggest.
It turns out that almost all of the bad news comes from two consecutive bad LFS numbers from November and December. As I discussed here, there's a certain amount of sampling error associated with the LFS data, and the standard errors are larger for subsamples. There's still a non-negligible probability that I'm making a big deal out of statistical noise. And there's as yet no other independent source that could confirm or contradict the LFS.
This is troublesome, because employment is generally a lagging indicator for the business cycle. If training new workers is costly, then firms would generally prefer not to lay off the workers they've already trained, and they will hesitate to hire until they are certain that they will be able to recover the costs of training new workers. So it just doesn't seem right that the first sign of a possible Quebec recession should be a drop in employment. Other things should have happened first.
But the LFS is all we have to work with, so that's what this post is based on.
Here is my rough-and-ready proxy for job creation: the number of people who have been at they current employment for less than three months (see also here). This is an important series because recessions are generally associated with a fall in job creation; job separations are generally stable.
You can see the drop in the early months of 2009, but not in the most recent data. Then again, given the delay with which a slowdown in job creation would show up in the short-tenure numbers - up to three months - the fact that nothing yet has happened in these data doesn't tell us much.
Here is my rough-and-ready proxy for job losses:
Here we definitely see something happening in November and December. It would be nice to cross-reference these numbers with the data for initial EI claims, but, alas, the most recent data are for October:
You can safely assume that that I will be looking at next Thursday's EI release with great interest.
The bad news is in the LFS numbers from November and December, and so far, the *only* available data for those months are from the LFS. They may be a glitch. If they aren't, we'll have to wait at least another month or two before we can get a better understanding of what's going on in Quebec - and why it *isn't* happening in the other provinces.
If this is due to a strike, as some other posters suggested, then the numbers won't show up in the EI claims. EI does not pay benefits for strikes. Further there is the significant number of people in non-standard employment that is uninsured or who have insufficient hours to make a claim.
Posted by: Determinant | January 12, 2012 at 10:44 PM
In the preceding thread on this topic, a commenter pointed to some statscan data that suggests the job losses a pretty widespread.
Could internal migration be part of it? Maybe they got *fed* up and moved to Canada.
Posted by: Patrick | January 13, 2012 at 07:51 AM
Since when does moving out of Quebec constitute moving to Canada? I thought that expression was confined to cranky old Newfoundlanders who refused to accept the New Order. ;)
Seriously, due to language preferences I believe Quebec workers are the least prone to move to another province of Canadian workers. If you want to work in French or can only work that language, Quebec is the place for you, maybe Moncton but that's it.
Posted by: Determinant | January 13, 2012 at 03:43 PM
Dunno about that. AB needs workers and if you're English is so-so, that's not too big a problem. I have colleagues from Asia whose English is honestly pretty bad, but we still manage pretty well. I hear quite a bit of French in these parts. Believe it or not, there's a Franco-Albertan community in northern AB (look at a map: St. Albert, Lac la Biche, Vermillion), and St. Albert is quite the upscale are (little to suburban for my tastes, but the school are apparently pretty good). You can send you kids to French school in Edmonton and St. Albert, live mostly in French, pay AB taxes, and never hear about separation again. What's not to like?
Posted by: Patrick | January 13, 2012 at 04:43 PM
Hehe. No Seperation? Jacques' head would explode. ;)
Posted by: Determinant | January 13, 2012 at 11:36 PM
Hold it, isn't there an Alberta separatist movement?
Posted by: Jim Rootham | January 15, 2012 at 11:56 PM
Could the fact that the LFS is reconciled with the 2006 census be an issue? It's now over 5 years since that census was taken. If there have been some notable population shifts since then, is it possible that the weight given to some LFS answers is wrong? And therefore skewing the results?
I'm not a statistician, but would love to hear the opinion of a statistician on this.
I'm grappling with a similar puzzle with the LFS at a more micro level in Calgary. It's not showing as many jobs being created as other evidence suggests has occurred.
Posted by: Wendy | January 16, 2012 at 02:57 PM
I was wondering about that. Of course, even if the 2011 census were usable, they'd still be using the 2006 numbers at this point.
But I wonder if there are other examples of wonky LFS numbers five years or so after the census that's used to balance the panel.
Posted by: Stephen Gordon | January 16, 2012 at 06:07 PM
Guess we need a statistician.
On the Quebec question, I do have the latest LFS for Montreal by NAICS code. Net, Montreal is down 20,000 jobs since January 2011. However, there are gainers and losers in the different industry categories. Biggest drop over the past 12 months has been FIRE (-20,000 jobs), followed by health care (-14,000 jobs) and info-culture (-10,000). Biggest gains in Educational services and Accommodation/food.
Does this help you figure out what is going on in any way?
Posted by: Wendy | January 17, 2012 at 05:33 PM
Wendy : given the size of these sectors, the data makes no sense. To lose 14,000 in health services requires massive cuts in hospitals, which didn't happened.
10,000 in communications cultures? Essentially everyone in the sector would be out in the street.
We need a very good statistician.
Posted by: Jacques René Giguère | January 17, 2012 at 06:58 PM
Jacques,
I agree that some of the stats don't make sense. In Calgary the stats also show a strong drop in FIRE employment which no one can find evidence of (and Calgary is a small enough place that a drop of 11,000 positions over the past 3 years would be noticed (7,000 in the past year alone).
I wonder if the LFS is up to something. Either a different methdology or weighting; FIRE jobs in other cities don't quite make sense to me either (and I've been tracking these closely since 2008, and the past few months is the first time things looked really odd).
We need a LFS statistician, I think. Not just any Statistician.
Posted by: Wendy | January 17, 2012 at 07:10 PM
Wendy: in the meanwhile, the press is going its usually abject ways.
Just look at this column by La Presse's Alain Dubuc ( who is an economist by trainig but turnedcoat for the media ethos, such at it is).
http://www.cyberpresse.ca/chroniqueurs/alain-dubuc/201201/10/01-4484578-le-party-est-fini.php
I didn't have the heart in me to check the anglo press.
Why,oh why, can't we have a better press corps?
http://www.google.com/search?hl=en&q=why+oh+why+can%27t+we+have+better+press+corps
Posted by: Jacques René Giguère | January 18, 2012 at 02:10 PM